In today's rapidly evolving business environment, the availability of funding has greatly diminished during the past few years, while the energy consumption and carbon emissions from IT have greatly increased. Therefore, understanding and optimizing both the financial and ecological impacts of IT infrastructure is crucial.

This article will guide you through the essentials of FinOps and GreenOps, two practices designed to enhance financial efficiency and sustainability in cloud and IT operations. We will also elaborate how both of these practices work towards a common goal.

Need for tighter control over cloud costs

Over the past couple of years, we have seen a great amount of investments in tech pulled away. This has been attributed to economic uncertainty, rise in interest rates, and ongoing wars. One factor is also the post-COVID investment boom slowing down, and the market correcting itself as a result.

Meanwhile, recent surveys show that a significant amount of organizations estimate they waste 40% of their cloud computing spending. Several organizations have even left or are considering leaving the cloud behind entirely, citing high cloud costs as the primary reason.

These recent trends have resulted in many companies looking into optimizing their spending in various ways, including reducing their spending on cloud platforms. Around the same time, FinOps started gaining momentum as a practice to help companies control their cloud spending.

FinOps, as explained by its principles

FinOps is a framework that goes beyond cloud cost optimization and considers how to maximize the business value of the cloud. Even though FinOps is not solely about cost optimization, several companies have managed to cut millions of dollars in cloud spending by including FinOps practices as part of their software development and delivery workflows.

In some shape or form, FinOps has been around since the early days of the cloud, and it gained a name around 2018 or 2019. Today, it is backed by a foundation of its own, the FinOps Foundation.

One way to examine FinOps is through its six principles.

Cross-team collaboration

The first thing to do in FinOps is to bring the right people together and not keep cloud cost optimization only as an engineering or finance project.

Specifically, you want the engineering, finance, product, procurement, and leadership teams to talk to each other to understand how the cloud costs relate to the overall company costs. From there, you can begin to understand how relevant cloud cost optimization is for you.

If this sounds familiar to you, this is because this is pretty much the first thing most software engineering methods suggest you do first including Agile, DevOps, SRE, and now also FinOps.

Cloud usage ownership

In many modern tech organizations, engineering teams are given autonomy on which tools to use, how to arrange their ways of working, how to build things, and even what kind of things to build. At the same time, they are also responsible for being able to deliver their software and maintain its quality, reliability, and security.

In the cloud, in many cases, engineering decisions affect the cloud costs directly, which means that engineers are almost always making the purchase decisions, whether consciously or not.

It then naturally makes sense that teams should also be responsible for cloud spending, as well.

Near real-time cost data

If you want teams to take ownership of cloud costs, they need to know what exactly is eating their cloud budget to evaluate cost optimization targets.

A key part of timely cost optimization is to have good cost reporting. The more real-time, the better. Quick feedback loops on spending encourage continuous improvements.

Value-driven decisions

Cost data on its own is not enough to make decisions on when to optimize spending. You also need to know how the cloud costs relate to the profit it generates for your business to ensure that effort is not wasted on minimal gains. For example, if you have very high cloud spending but also very high revenue that far surpasses the costs, then cost optimization might not make sense.

Leveraging variable cost model

From traditional data center environments, we are used to the idea that compute capacity is fixed for the long term. It takes several months to deliver the hardware and put it into use.

In the cloud, you can get additional capacity in minutes and sometimes in seconds and only pay for the time you use it.

Part of FinOps is to take advantage of this model and make purchase decisions as you go to minimize cloud costs. This is often achieved by designing software to take advantage of the model and automating the purchase decisions based on observed utilisation.

Facilitating team

Finally, each organisation is suggested to have a team to facilitate all of the FinOps principles mentioned above. The role of this team is to oversee cloud utilization across the whole organisation, help find optimisation opportunities, spread knowledge on managing costs as part of the team development workflow, and help negotiate rates with the cloud vendors.

The team can be built in several ways depending on the size of the organisation: Larger organisations may want to hire dedicated personnel for the team, while smaller organisations may want to create task forces among existing personnel.

The carbon footprint of IT is growing

As the world becomes increasingly digital and innovations such as new AI models are produced, the demand for energy and natural resources in the IT sector also increases. It is estimated that IT consumes one tenth of the world’s electricity. In addition to electricity, computer hardware also consumes water for cooling to prevent hardware from overheating. Moreover, production of new hardware consumes a vast amount of minerals and energy during the hardware production process and delivery.

With the resource usage, the amount of carbon emissions also rises rapidly. It is estimated that IT contributes 5-10% of annual global carbon emissions, which is more than the airline industry.

At the same time, data centers are largely underutilized. It is estimated that only 10-20% of the on-premise data center capacity and 65% of the cloud infrastructure capacity is utilized. This means that a lot of the energy used and emissions produced are not spent on producing value.

Regulations in the field are catching up: The recent EU directive for corporate sustainability reporting (EU CSRD) mandates that companies must adopt energy-efficient practices. In the IT sector, this means that organizations must focus on improving their hardware capacity utilization and choosing more power efficient solutions in both hardware and software.

Even though the sustainability reporting directive primarily targets large organizations, the smaller companies that are part of the supply chain for the larger organizations will need to participate in the same reporting directive one way or the other. Therefore, companies that adopt a sustainability strategy will have a competitive advantage over those that haven’t as they are more likely to be selected as suppliers for larger organizations.

Reducing the energy usage and carbon emissions with GreenOps

To combat the energy waste and increase in carbon emissions, we can use GreenOps to optimize our software infrastructure.

At its heart, GreenOps is about minimizing the ecological impact of software infrastructure and business operations. There are three ways GreenOps works: energy efficiency, resource management, and sustainable sourcing.

Energy efficiency

Energy efficiency is about optimizing software and hardware to consume less energy while completing the same amount or more work.

On the infrastructure side, we can consider using hardware that consumes less energy for the same computing needs.

On the software side, we can optimize the software component efficiency (whether they are in-house built or platform provided) such that less time is spent on performing the tasks, and thus reduce the overall energy consumption that way.

Resource management

Another way to improve sustainability is to manage our existing resources efficiently.

By recycling hardware, we take full advantage of the energy that was spent on producing the hardware. This can be achieved by designing software such that it works efficiently enough on older hardware. Alternatively, old hardware from production environments can be reused in less mission-critical environments such as development and test environments.

A large part of resource management is also to eliminate waste. For example, stopping unused cloud resources, stopping tasks that produce unused results, and deleting unused data to reduce the need for more storage hardware.

Sustainable sourcing

When it comes to buying new hardware or software, we can evaluate how the suppliers take sustainability into account. Are they transparent about their carbon footprint? Do they take measures to lower their energy consumption when providing their services?

Similarly, we can take into account how clean the energy sources are for the hardware and software we run. In other words, how much renewable energy is used for powering what we do? In some cases, we can even shift the computations to occur only when there is more clean energy available, and thus reduce the overall consumption of non-renewable energy while performing the same amount of work.

Green Lean: Where FinOps and GreenOps meet

Now that we are familiar with FinOps and GreenOps, you might be wondering how they are related to each other.

Our answer is very simple: By optimising the utilisation of compute resources, we need less compute resources, which improves sustainability, but also benefits us financially.

To achieve this, we can leverage the principles from FinOps in GreenOps such as near real-time reports, and placing resource utilisation in the business context.

Besides GreenOps and FinOps, we have noticed that we can use Green Code and Performance Engineering practices to contribute to sustainability. We call the combination of these practices Green Lean.

GreenOps: Minimising ecological impact of infrastructure
FinOps: Maximising cloud business-value
Green Code: Optimising software energy efficiency
Performance Engineering: Optimising software and infrastructure performance

All that said, we do see challenges around the corner. According to Jevon's paradox, efficiency increases overall demand higher than the original utilisation. For example, if we are to reduce the cloud costs, there is a chance that the customer will use the extra money on other new cloud resources.

Nevertheless, we see that we need to keep optimizing to at least eliminate any wasteful computation. In addition to optimization, we also contribute back to nature by sharing 10% of our profits.

At Kanto, we are experts in both FinOps and GreenOps. We provide short-form assessments and hands-on engineering to our customers to help them get the most out of their cloud systems both financially and ecologically. If you’re looking for experts to help you take the first steps in FinOps and/or GreenOps, we’ll be happy to help you more. Send us a message at hello@kantocompany.com, and we’ll tell you more.

FinOps + GreenOps - Unlocking Financial and Environmental Benefits